See discussions, stats, and author profiles for this publication at: https://www.researchgate.net/publication/280041888
Adhesion Testing of Sealants
Chapter · August 2009
DOI: 10.1201/9781420008630-c7
CITATION
READS
1
7,534
2 authors, including:
Voytek Gutowski
University of Melbourne
263 PUBLICATIONS 1,283 CITATIONS
SEE PROFILE
Some of the authors of this publication are also working on these related projects:
Electrostatic coating of non-conductive substrates View project
Solid wood adhesion View project
All content following this page was uploaded by Voytek Gutowski on 30 September 2015.
The user has requested enhancement of the downloaded file.
7
Adhesion Testing
of Sealants
W (Voytek). S. Gutowski and A.P. Cerra
Contents
7.1 Introduction.................................................................................................... 192
7.2 Typical Function and Geometry of the Sealant Bead in a Joint..................... 193
7.3 Degradation Factors and Stresses Imposed on the Sealants during
Service Life.................................................................................................... 193
7.4 Principles for Assessing the Adhesion of Sealants........................................ 194
7.5 Stress at the Sealant–Substrate Interface....................................................... 195
7.5.1 Maximizing Stress Along the Interface of Peel Test Specimens....... 196
7.5.1.1 Mechanics of Peel Test....................................................... 196
7.5.1.2 Relationship between the Elasticity Modulus and
Sealant Thickness................................................................ 201
7.5.1.3 Effective Stress Control at the Sealant–Substrate
Interface in Peel Specimens................................................ 201
7.5.2 Joint Geometry Parameters Controlling Stress at the Sealant–
Substrate Interface under Tensile Strain............................................ 201
7.5.2.1 Alternative Utilization of the Joint Geometry Design........ 201
7.5.2.2 Dependence of Elasticity Modulus on the Sealant
Dimensions in Joints Subjected to Static or Dynamic
Tensile Stresses...................................................................202
7.6 Other Factors Influencing the Mechanical and Rheological Properties of
Sealants..........................................................................................................205
7.6.1 Strain Rate and Temperature..............................................................205
7.6.1.1 Strain Rate in Conventional Sealant Joints.........................205
7.6.1.2 The Influence of Strain Rate and Temperature on the
Tensile Strength of Sealants................................................205
7.7 Test Methods for Determining Adhesion of Sealants....................................206
7.7.1 Peel Test Methods..............................................................................206
7.7.2 Test Methods based on Tensile Specimens........................................ 211
7.7.2.1 Deficiencies in Existing Test Methods Involving
Tensile Specimens............................................................... 211
7.7.2.2 CSIRO Test Method for Determining Adhesion of
Sealants............................................................................... 212
7.8 Contemporary Standards For Determining Sealant Performance.................. 217
7.8.1 RILEM Accelerated Durability Tests................................................ 218
191
9162_C007.indd 191
3/17/09 8:27:18 PM
192
Handbook of Sealant Technology
7.8.1.1 Static Conditioning Protocol............................................... 218
7.8.1.2 Dynamic Conditioning Protocol......................................... 218
7.8.2 RILEM Protocols Involving Artificial Weathering............................ 219
7.8.3 \Thermomechanical Cycling and Determination of Sealant
Durability........................................................................................... 219
7.8.4 Most Recent RILEM Recommendations for Sealant Durability
Assessment........................................................................................ 220
7.9 Summary........................................................................................................ 221
References............................................................................................................... 221
7.1
Introduction
Elastomeric sealants are an important class of specialty adhesives used by the building, automotive, aerospace, electronics, and other industries for joining, gap filling
and, frequently, structural bonding of a variety of engineering materials. To fulfill its
designated functions, the sealant needs to retain cohesive integrity as well as adhere
to the joined substrates throughout the lifetime of the sealed or bonded structure.
During their service, sealants are typically exposed to a range of static and
dynamic stresses imposed by external loads; cyclic compressive–tensile forces arising from the structure’s movements and deformations; and also caused by environmental factors such as solar radiation, temperature fluctuations, moisture ingress,
and others.
A combination of these factors frequently leads to changes in the material’s internal structure and properties such as crosslink density and chain mobility, and leads
to changes in its rheological and mechanical properties. The ensuing increase in the
elasticity modulus and the reduction of the material’s elongation capability may lead
to the increase of stresses within the sealant bead and, more importantly, at the sealant–substrate interface. Once stresses exceed the strength of interfacial or internal
bonds, cracks are initiated. The onset of crack formation in the continuous presence
of cyclic joint movement and mechanical and environmental stresses is typically
followed by gradual crack propagation along the path of the weakest element of the
joint. In most cases, this is the sealant–substrate interface.
To withstand the detrimental influence of stresses and degradation-inducing factors, sealants are formulated to exhibit the following long-term properties:
• Reliable and stable adhesion to a variety of engineering substrates
• Stable bulk properties, for example, cohesive strength, modulus of elasticity,
and elongation capability, which are not adversely affected by ultraviolet or
infrared radiation from solar exposure or by other environmental factors,
and which hence allow accommodation of all the thermal and structural
movements of the bonded elements
This chapter addresses the following fundamental and engineering issues concerning in-service performance of elastomeric sealants with emphasis placed on the
sealant–substrate adhesion:
9162_C007.indd 192
3/17/09 8:27:18 PM
Adhesion Testing of Sealants
193
1. Stresses and degradation factors during the service life of sealants
2. Fundamental factors influencing stresses at the sealant–substrate interface,
and adhesion of sealants
3. Methods for determining the strength of adhesion between the sealant and
adjacent substrates
4. Industry standards and protocols for determining sealant performance
The experimental data discussed, and the ensuing performance–property relationships developed in this chapter are the result of an extensive research program carried out by the CSIRO research team in collaboration with key suppliers of structural
and nonstructural elastomeric sealants such as silicones, polyurethanes, polysulfides,
epoxy–polyurethane hybrids, and others.
7.2 Typical Function and Geometry of
the Sealant Bead in a Joint
In typical construction applications, sealants are used as weather seals to prevent the
ingress of water, air, and other undesirable elements. In another key role as structural
adhesives, they are designed to permanently attach panes of glass or cladding panels
to the structural members of a façade. Figure 7.1 illustrates examples of joint designs
for these two typical applications of sealants.
7.3 Degradation Factors and Stresses Imposed
on the Sealants during Service Life
Throughout the service life of a sealed or bonded structure, the sealant and its interfaces with adjacent substrates are continually subjected to a set of complex and concurrently imposed environmental and mechanical stresses and degradation factors,
such as the following:
• Load-controlled stress, imposed by external loads and forces, for example,
deadloads, wind pressure, and structural load deformation.
• Displacement-controlled stress, caused by cyclic or permanent variations
in the joint width. These are typically imposed by thermal movements of
adjacent cladding panels, for example, in building facades or the external
body panels of vehicles such as buses, trams, and boats.
• Hydrothermal stresses, caused by the ingress of water molecules (liquid
water or vapors) to the sealant–substrate interface and by increase of reaction rate (e.g., hydrolytic chain scission) in response to increased ambient or
service temperature.
• Photochemical factors, for example, photoinduced chain scission or crosslinking, and catalytic reduction of activation energy of reactions at the sealant–substrate interface.
9162_C007.indd 193
3/17/09 8:27:18 PM
194
Handbook of Sealant Technology
Weather
sealant
Glass
panel
Structural
sealant
Glass
panel
Structural
sealant
A
D
D
Figure 7.1 Typical designs of joints with elastomeric sealants: (a) structural sealant in a
joint, also performing the role of a weather seal [1]; (b) weather seal A between façade panels
D. (Adapted from American Society for Testing and Materials (ASTM) Standard No C119305: “Standard guide for use of joint sealants,” 2005.)
The loss of adhesion through delamination between the sealant bead and the substrate occurs when the fracture energy of the interphase, comprising an array
of molecular bridges across the sealant–substrate interface, is lower than the
fracture energy of the bulk sealant material. Hence, the configuration of the test
specimen must be selected and optimized to maximize the stress at the interface,
while other test conditions such as strain or deformation rate and environmental
stress factors mimic realistic conditions imposed on the actual joint during its
service life.
7.4 Principles for Assessing the Adhesion of Sealants
As outlined in Section 7.2, in most typical applications, sealants perform the role of
structural adhesives or weather seals preventing the ingress of liquid water or vapors,
9162_C007.indd 194
3/17/09 8:27:19 PM
195
Adhesion Testing of Sealants
air, and other undesirable elements to the sealed structure. The following are the key
factors essential for the satisfactory performance of sealants in any designated role:
1. Retention of cohesive integrity of bulk sealant material
2. Retention of adhesion between the sealant and substrate material
Most of the current standardized approaches for determining sealant performance
require demonstration of cohesive integrity of the sealant after exposing it sequentially
or concurrently to a number of static or cyclic degradation factors and stresses. While
the foregoing criterion (retention of cohesion) is frequently satisfied, it is the strength
of the array of chemical bonds across the sealant–substrate interface that plays an
equal or more important role for ascertaining satisfactory joint performance.
To demonstrate satisfactory adhesion, one needs to subject the interface to a set of
conditions likely to promote disruption of interfacial bonds before the cohesive failure
of bulk sealant may occur. To prevent the latter from occurring first, the task needs to
be accomplished by the concurrent use of degrading conditions together with mechanical stress of the magnitude below the cohesive strength of the sealant material.
The following are the most significant difficulties in accomplishing this demanding task:
1. How to reliably detect substandard adhesion at the sealant–substrate interface when strength and environmental stability can be masked by the elastomeric nature and premature cohesive failure of the sealant material.
2. How the joint dimensions, loading conditions, and other factors influence the
physical properties of the sealant and the magnitude of interfacial stress.
3. How the results relate to the long-term durability of the sealant in service.
4. How to simulate a realistic set of degradation factors most likely to: (i)
induce sealant aging, and (ii) promote adhesion failure, in order to reliably
predict long-term adhesion and cohesion durability.
7.5 Stress at the Sealant–Substrate Interface
In accordance with the principles of continuum mechanics, for sealant deformations
to remain within the region of the material’s elastic deformation, the magnitude of
stress, including that at the interface, is directly related to the material’s properties
such as modulus of elasticity ([E]: Young’s modulus) and strain [ε] through the following relationship:
E=
stress (σ)
strain (ε)
(7.1)
Equation 7.1 uses an average stress value to determine the criterion for joint failure and completely ignores stress concentration near the edges of the stretched
sealant or other loci resulting from intrinsic material defects such as air bubbles
or cracks.
9162_C007.indd 195
3/17/09 8:27:19 PM
196
Handbook of Sealant Technology
Tensile Stress (MPa)
Stress at failure
∆L
L
E = tanα
α
Extension at failure
Strain, ε = (∆L/L · 100)
Figure 7.2 Typical stress–strain curve and deformation of a viscoelastic sealant material
in a joint subjected to tensile force (L is the initial width of the sealant bead, ΔL is the sealant
elongation, and E is the Young’s modulus of the sealant).
In practice, due to the sealant’s rheological properties leading to bead deformation with increasing strain, as schematically illustrated in Figure 7.2, edge effects
occur with stress concentrating at the edge of the joint. The latter typically becomes
the point (or line) where the crack initiates before propagating along the interface
(in the case of poor adhesion) or veering into the bulk sealant when adhesion is
satisfactory and cohesive strength controls the overall joint performance.
The stress peak value at the edge of the deformed joint, which can be computed
using the finite elements method (FEM) or finite differences analysis (FDA) exceeds
the average stress estimated through Equation 7.1. This difference increases nonlinearly with increasing deformation.
It is noticeable from Equation 7.1 that, for a constant deformation, that is, strain (ε)
exerted on the sealant, the resultant average stress (σ) at the sealant–substrate interface increases with increase in the modulus of elasticity (E). This fact (the increase
of interfacial stress with the modulus increase) is practically utilized in our test
methods for determining the strength and quality of sealant adhesion. It is comprehensively demonstrated in Sections 7.5.1 and 7.7.1.
7.5.1
Maximizing Stress Along the Interface of Peel Test Specimens
7.5.1.1 Mechanics of Peel Test
One of the most common techniques for testing sealant adhesion is the peel test (see
Figure 7.3), the principles of which are discussed in detail in References 3–6.
9162_C007.indd 196
3/17/09 8:27:20 PM
197
Adhesion Testing of Sealants
Reinforcing mesh
Sealant
σ11
ψ
(a) Peel specimen
Tensile stress
Theoretical
Experimental
(b) Stress distribution
(x)
Compressive
stress
Figure 7.3 Distribution of tensile stresses σ11 at the sealant–substrate interface ahead of
the advancing peel front (x). ψ is the peel angle. (From D.H. Kaelble, Trans. Soc. Rheology,
4, 45, 1960.)
According to Kaelble [3], the distribution of tensile stress σ11 at a distance x from
the peel front is expressed by the following relationship:
where
σ11 = σ 0 (cosβpx + κ sin βpx) exp(βpx),
βp = (
Eab 1/4
)
4EIha
κ = (βpm/βpM + sin ψ),
(7.2)
(7.3)
(7.4)
where σ 0 is tensile stress at x = 0; Ea is tensile modulus of the sealant material; E is
tensile modulus of the flexible backing mesh material; ha is sealant layer thickness
between the substrate and flexible backing mesh material; I is moment of inertia of
the peeled sealant strip; M is moment of the peel force; ψ is the peel angle.
According to Kaelble’s theory [3–5] the peel force [F] per unit width [b] of the
specimen is given by
F
ha κσ 2
=
b 2 E (1 - cos ψ )
9162_C007.indd 197
(7.5)
3/17/09 8:27:22 PM
198
Handbook of Sealant Technology
A drawback of Kaelble’s equation (Equation 7.5) is that it does not consider the
influence on the peel force of the following factors: (1) strain rate (peel-off speed),
and (2) variability of elasticity modulus with the sealant geometry (e.g., sealant
layer thickness).
To close this gap, experiments were carried out in CSIRO laboratories to investigate the influence of strain rate and sealant bead geometry (controlled through
varying the sealant thickness) on the magnitude of resultant peel force. The strain
rate was controlled within the range of 0.05 to 100 mm/min (0.05; 0.5; 1; 2.5; 5; 10;
25; 50; 100 mm/min) while sealant thickness was varied from 0.1 to 0.6 mm. Graphs
in Figure 7.4 illustrate the outcome of these experiments. The influence of sealant
bead thickness on the elasticity modulus was also investigated and is illustrated in
Figure 7.5.
An analysis of Equation 7.5 from the perspective of experimental data illustrated
in Figures 7.4 and 7.6 leads to a new insight into the mechanism of stress control at
the peel front of the sealant–substrate interface.
It is noticeable from Figure 7.4 that an increase in the sealant layer thickness
between the rigid and flexible substrates results in an increase in the peel force, as
earlier reported by Gent and Hamed [6]. The graphs in Figure 7.4 also demonstrate
that the peel force (and hence fracture energy) strongly depends on the peel rate (ε˙).
Hence, Kaelble’s equation (Equation 7.5) requires correction for the peel rate, that is,
F
ha κσ 2 . f (ε˙)
=
b 2 E (1 - cos ψ )
(7.6)
50
0.6 mm
Peel Force (N/25 mm)
40
0.4 mm
30
0.3 mm
0.2 mm
0.1 mm
20
Sealant thickness
10
0
0
20
40
60
80
·
Peel Rate, ε(mm/min)
100
120
Figure 7.4 The influence of peel rate and sealant bead thickness on the magnitude of peel
force (sealant type: Dow Corning 795; temperature: 20°C; failure mode in all cases: 100%
cohesive failure within the sealant).
9162_C007.indd 198
3/17/09 8:27:25 PM
199
Adhesion Testing of Sealants
Young’s Modulus, E (MPa)
3.0
2.0
1.0
0
1.0
2.0
3.0
Sealant Bead Thickness (mm)
Figure 7.5 The relationship between sealant (Dow Corning 983) bead thickness and modulus of elasticity (Young’s modulus) at 20°C. (From W.S. Gutowski, L. Russell, and A. Cerra,
in Science and Technology of Building Seals, Sealants, Glazing and Waterproofing, Volume
2, ASTM STP 1200, J.M. Klosowski (Ed.), p. 87, American Society for Testing and Materials,
Philadelphia, PA, 1992.)
It follows from Equation 7.4 that κ = 1 for the peel angle ψ = 180° used in our experiments. Considering this and using Equation 7.6, the magnitude of tensile stress σ at
the peel front can now be determined as
 FE
1 

σ=
i
f (ε ) 
 bha

1
2
(7.7)
Equation 7.7 demonstrates that, for a constant specimen width b, the tensile stress at
the peel front is proportional to the experimentally determined peel stress F/b and
the sealant’s elasticity modulus E, and inversely proportional to both the thickness of
the sealant layer ha and the function f(ε˙), which describes the relationship between
the peel force F and peel rate ε˙. It can also be seen from Equation 7.7 that the stress
at the peel front increases in direct proportion to the reduction in the thickness of the
sealant layer between the rigid and flexible substrates. It means, for example, that a
fivefold reduction in the sealant thickness increases the stress level at the peel front
five times in comparison to that achievable at the original thickness.
As discussed in detail References in 7–10, the peeling of a flexible layer of sealant
(backed by an elastic backing tape) from a rigid substrate at a constant peel force F
and a peel angle ψ involves energy changes as the crack propagates by an increment
Δx through the sealant or along the interface. These energy changes must be balanced in accordance with the energy conservation principle, that is,
W+Γ+V=0
9162_C007.indd 199
(7.8)
3/17/09 8:27:25 PM
200
Handbook of Sealant Technology
where W is the work of peeling, Γ is the energy required to create a new fracture surface, and V is the peel deformation energy (comprising stored elastic energy and the
work associated with deforming the sealant along the path Δx) for a newly created
interfacial area A = [(b) · Δx] where b is the width of the peel specimen.
The values of the parameters in Equation 7.8 are determined as follows:
W = F (λ – cos ψ) Δx
(7.9a)
Γ = − G · b · Δx
(7.9b)
V = (Saha + Sh) bΔx
(7.9c)
where F is the measured peel force, λ is the extension rate of the flexible backing tape,
ψ is the peel angle, b is the peel specimen width, ha is the thickness of sealant layer
between the flexible and rigid substrates, h is the thickness of the flexible backing tape,
Sa is the deformation energy per unit volume of sealant, S is the deformation energy
per unit volume of backing tape, and G is fracture energy per unit interfacial area.
The total fracture energy G per unit interfacial area in Equation 7.9b comprises
the following two components [9]:
G = Go + Uha
(7.10)
where Go is the intrinsic interfacial fracture energy equal to the actual magnitude of
intermolecular forces at the sealant–substrate interface, and Uha is the plastic dissipation energy irreversibly lost by yielding in the zone around the deformed tip of the
propagating crack (U = dissipation energy per unit volume of sealant, ha = thickness
of the sealant layer between the flexible and rigid substrates).
Substitution of Equations 7.9a and 7.9c into Equation 7.8 gives the general formula [10] for fracture energy (determined from peel test), which is valid for any
viscoelastic adhesive such as an elastomeric sealant:
G=
F
b
(λ – cos ψ) + Saha + Sh
(7.11)
If the flexible backing tape can be considered inextensible at the prevailing peel
forces, then λ = 1 and S = 0. For this case, Equation 7.11 simplifies as follows:
F
G = (1– cos ψ ) + Sa ha
b
(7.12)
For a slightly extensible backing tape, Equation 7.12 becomes
F 1+λ
G= (
– cos ψ ) + Sa ha
b 2
9162_C007.indd 200
(7.13)
3/17/09 8:27:29 PM
Adhesion Testing of Sealants
201
Equations 7.12 and 7.13 can be used for determining the fracture energy from peel
tests carried out at a constant peel rate [ε˙] and any peel angle.
It is clear from Equations 7.12 and 7.13 that the adhesion forces exerted across the
sealant–substrate interface (and related to fracture energy G of the interface through
these equations) can be most accurately determined when the plastic dissipation
energy is minimized through the reduction of Sa and ha terms. This is effectively
achieved by minimizing the thickness of the sealant layer, which, in turn, provides
the desired reduction in the deformation energy per unit volume of sealant due to the
increase in the material’s modulus of elasticity with decreasing thickness, as shown
previously. The reduction of peel angle ψ leads to a further reduction in the plastic
deformation losses.
7.5.1.2 Relationship between the Elasticity Modulus and Sealant Thickness
It has been demonstrated earlier [11] that the apparent modulus of elasticity of elastomeric sealants strongly depends on the thickness of the material bonded between
rigid substrates. Figure 7.5 illustrates an example of this relationship for Dow Corning
two-component DC 983 sealant.
It shows that the modulus of elasticity of the sealant remains constant when the
thickness of the sealant layer is reduced from 3 mm to 1 mm, but subsequently starts
to increase exponentially when the thickness becomes less than approximately 0.5
mm. At 0.25 mm, the modulus is double that at 2 mm.
7.5.1.3 Effective Stress Control at the Sealant–
Substrate Interface in Peel Specimens
Considering the experimental data on peel force and elasticity modulus presented
in Figures 7.4 and 7.5, respectively, the magnitude of the stress at the peel front in
relation to the thickness of the sealant bead can be determined from Equation 7.7, as
illustrated in Figure 7.6.
It can be concluded from the foregoing analysis of the theory and experimental data
that the magnitude of the tensile stress at the sealant–substrate interface can be effectively
maximized by reducing the thickness of the sealant layer between the bonded substrates.
This result has been successfully applied in the experimental design of the specimen
geometry and test protocol for preferentially assessing the adhesion between the sealant
and substrate rather than the cohesive strength. It is discussed in detail in Section 7.7.1.
7.5.2 Joint Geometry Parameters Controlling Stress at the
Sealant–Substrate Interface under Tensile Strain
7.5.2.1 Alternative Utilization of the Joint Geometry Design
The average tensile stress σ at the sealant–substrate interface depends, according to
Equation 7.1, on the strain ε, and increases with the sealant’s modulus of elasticity E.
A detailed analysis of the influence of joint geometry on the behavior of a sealant
within the joint can be effectively utilized as follows:
9162_C007.indd 201
3/17/09 8:27:29 PM
202
Stress at Sealant/Substrate Interface, (σ)
Young’s Modulus, E (MPa) and Peel Force, F/b (N/mm)
Handbook of Sealant Technology
6
5
Stress at sealant/substrate
interface [Eqn.(7)]
4
3
Peel force (experimental)
2
Young’s modulus
(experimental)
1
0
0
1.0
2.0
3.0
Sealant Bead Thickness (mm)
Figure 7.6 Tensile stress (σ) at the peel front at the sealant–substrate interface as a function of sealant bead thickness in accordance with Equation 7.7. (From W.S. Gutowski, L.
Russell, and A. Cerra, in Science and Technology of Building Seals, Sealants, Glazing and
Waterproofing, Volume 2, ASTM STP 1200, J.M. Klosowski (Ed.), p. 87, American Society
for Testing and Materials, Philadelphia, PA, 1992.)
• In the engineering design of sealed joints to minimize the stress in structurally bonded systems and to ensure adequate movement capability when
sealants are applied as weather seals, or
• In the design of test joints to maximize the stress at the sealant–substrate
interface, thereby preferentially generating interfacial failure
7.5.2.2 Dependence of Elasticity Modulus on the Sealant Dimensions
in Joints Subjected to Static or Dynamic Tensile Stresses
It has been demonstrated in Section 7.5.1.2 (see Figure 7.5) that the apparent elasticity modulus of electrometric sealants is inversely proportional to the joint “aspect
ratio” (ratio of depth to width) when subjected to peel forces. The question arises if a
similar relationship exists for tensile loading.
To answer this question and to address its implications for the design of test protocols for assessing adhesion strength when subjecting the sealant–substrate interface
to static or cyclic tensile stress, a set of experiments were carried out employing the
tensile joints schematically illustrated in Figure 7.7.
The joint geometry was controlled by varying the sealant bead dimensions within
the following range of parameters:
• Sealant bead thickness (G: glueline): 6–12 mm
• Sealant bead width (B: bite): 6–20 mm.
9162_C007.indd 202
3/17/09 8:27:29 PM
203
Adhesion Testing of Sealants
Glueline (G)
40 mm
Bite (B)
50 mm
Figure 7.7 Geometry of tensile specimens for determining the modulus of elasticity, tensile strength, and elongation at failure of structural and weather sealants. (Note: The terms
glueline and bite are commonly used by the building and construction industry instead of
those typically used in adhesion science, that is, G = bondline thickness or sealant bead thickness or glueline, and B = joint width or sealant bead width or bite.
The resultant joint aspect ratio [B/G], defined here as
Aspect ratio [AR] =
[B] bead width
[G] glueline thickness
was controlled within the range of [AR] = 0.67 to 3.33. This represents changing the
joint configuration from “oblong” to “flat,” the latter exhibiting a bead width 3.33
times greater that the bead thickness.
The results for modulus of elasticity E for a range of joint geometries as investigated in this experiment for a typical structural silicone are provided in Table 7.1.
Figure 7.8 depicts the relationship between the elasticity modulus E and relative thickness of sealant bead represented by the joint’s aspect ratio B/G. It can be
seen from the comparison of Figures 7.5 and 7.8 that the trend observed for the
sealant in tensile joints follows that earlier observed for the peel joint configuration, that is, that the material’s modulus of elasticity (in tensile mode) increases
with the reduction of the relative thickness of the sealant layer or increasing
aspect ratio.
This result can be used to design tensile test joints that preferentially fail interfacially rather than cohesively.
9162_C007.indd 203
3/17/09 8:27:30 PM
204
Handbook of Sealant Technology
Table 7.1
Modulus of Elasticity of a Typical Structural Silicone as a
Function of Varying the Joint Geometry
Joint Dimensions
Width [B] ×
Thickness [G]
(mm)
Aspect Tatio
(B:G)
8 × 12
8 × 10
6×6
8×8
12 × 12
12 × 10
8×6
16 × 12
12 × 8
16 × 10
20 × 12
12 × 6
16 × 8
20 × 10
16 × 6
20 × 6
0.67
0.80
1.00
1.00
1.00
1.20
1.32
1.33
1.50
1.60
1.67
2.00
2.00
2.00
2.66
3.33
Joint
Cross-sectional Area
(mm2)
Modulus of
Elasticity: E
(MPa)
96
80
36
64
144
120
48
192
96
160
260
72
128
200
96
120
0.75
0.66
0.82
0.89
0.87
0.95
0.80
1.10
1.25
1.15
1.06
1.45
1,26
1.34
1.27
1.41
Note: Test conditions: T = 20°C; strain rate = 50 mm/min.
Modulus (MPa)
1.6
Young’s modulus (E)
1.2
0.8
Shear modulus (G)
0.4
0
0.6
1.0
1.4
1.8
2.2
2.6
3.0
3.3
Aspect Ratio
Figure 7.8 The relationship between elasticity modulus E and relative thickness of sealant
bead represented by sealant bead aspect ratio (width:thickness).
9162_C007.indd 204
3/17/09 8:27:30 PM
205
Adhesion Testing of Sealants
7.6 Other Factors Influencing the Mechanical
and Rheological Properties of Sealants
7.6.1
Strain Rate and Temperature
7.6.1.1 Strain Rate in Conventional Sealant Joints
Sealants in applications such as building facades and other structures undergo cyclic
strain imposed by daily and seasonal fluctuations of ambient temperature, which
causes thermal movement in cladding panels. The observed sealant extension–compression cycling depends on the rate and extent of the temperature variation, substrate dimensions, and coefficient of thermal expansion. The typical range of strain
rates is as follows [12]:
• Daily cycling rate: 10 −1–10 −2 mm/min
• Seasonal cycling rate: 10 −5–10 −6 mm/min
Consequently, test protocols for determining sealants’ engineering properties,
including adhesion, must be determined at these strain rates and carried out within
the range of actual service temperatures.
7.6.1.2 The Influence of Strain Rate and Temperature
on the Tensile Strength of Sealants
It has been shown by Ferry [13] that the effects of strain rate and temperature on
the rheological and mechanical properties of a polymer, for example, sealant tensile
strength, are equivalent and thus can be superimposed by shifting each individual
temperature-related property curve to a collective one to give a single master curve.
The resultant master curve is described [13] by Equation. 7.14:
STo
•
ε TaT
∞
=
∫



•



M(τ ) • τ 1-e- ε/ε aT  • d ln τ
-∞
(7.14)
where S is the stress; ε˙ is the strain rate; T0 is arbitrary reference temperature,
for example, the glass transition temperature of the polymer; T is the actual test
temperature at which the stress–strain curve (S versus ε) is determined; τ is time;
M(τ) is relaxation distribution function; ε is strain; and a T is the temperature shift
factor.
The value of the temperature shift factor aT is estimated through an Arrhenius
approach using the Williams–Landel–Ferry (WLF) equation (Equation 7.15a):
log aT =
9162_C007.indd 205
C1 (T − T0 )
C2 + T − TS
(7.15a)
3/17/09 8:27:31 PM
206
Handbook of Sealant Technology
where T is the test temperature; T0 is the reference temperature, for example, the
glass transition temperature of the polymer; and C1, C2 are experimental constants.
According to Ferry [13], in his first application of the WLF equation, the average
values of C1 and C2 were obtained by fitting experimental data to a range of polymers and were estimated to be 17.44 and 51.6, respectively [14]. It was, however,
pointed out by Ferry that the actual variations of C1 and C2 from one polymer to
another were too great to permit the use of these constants as “universal’ values,
as is often done by many researchers. Ferry pointed out that, in a somewhat better
approximation, fixed values of C1 = 8.86 and C2 = 101.6 should be used in conjunction with a reference temperature T0, which was allowed to be an adjustable
parameter but generally fell about 50°C above the Tg, consequently giving rise to
Equation 7.15b:
log aT =
8.86 (T − Tg − 50)
101.6 (T − Tg − 50)
(7.15b)
The WLF formula (Equation 7.15b) gives satisfactory results over the temperature
range: Tg < T < Tg + 100°C.
An example of the master curve construction using the WLF approach is illustrated in Figure 7.9 for the tensile strength of joints made using polybutadiene-styrene elastomeric adhesive, PB-SR (Tg = – 40°C).
Once converted into a master curve, the rheological or mechanical properties of
the sealant can be easily estimated over the range of more than ten decades of the
reduced test rate (ε˙aT), which is easily converted into real time and strain rate.
Figure 7.10 illustrates the master curve, as developed by us earlier [12], representing the tensile strength of the Dow Corning 795 sealant, which in addition to
temperature T and strain rate ε˙, also considers cross-sectional area A of the sealant
bead in the joint.
7.7 Test Methods for Determining Adhesion of Sealants
7.7.1
Peel Test Methods
Historically, one of the most common tests for studying the adhesion of sealants has
been the peel test, commonly conducted at a peel angle of 180° and strain rate (peel
rate) of 50 mm/min (e.g., ASTM C-794 [16]). After appropriate artificial aging of the
specimen (water immersion, elevated temperature, freezing, etc.), the peel force and
failure mode are recorded. Although the specimen geometry does not realistically
resemble that of a typical joint, peel tests can reveal underlying adhesion problems if
the factors controlling the failure mode are well understood.
A thorough understanding of the mechanics of peel testing (see Section 7.5.1.1),
especially the use of appropriate specimen geometry (i.e., sealant bead thickness)
and test conditions is essential for promoting the fracture of interfacial bonds rather
than cohesive failure of the bulk sealant. This, of course, is essential for studying
adhesion behavior.
9162_C007.indd 206
3/17/09 8:27:32 PM
207
Adhesion Testing of Sealants
50
–30°C
–20°C
0°C
22°C
50°C
Tensile Strength [kg/cm2]
40
–30°C
–20°C
30
0°C
20
22°C
10
0
50°C
–4
–2
0
2
Strain Rate: Log ε· [m/s]
–4
4
0
8
· [m/s]
Reduced Rate: Log (aTε)
(a)
(b)
Figure 7.9 (a) Tensile strength versus test rate curves for joints comprising PET (Mylar)
substrate and PB-SR elastomeric adhesive tested at various temperatures: −30°C to +50°C,
(b) Master curve for the tensile strength of joints tested in (a) versus reduced test rate (ε˙ aT).
(From A.N. Gent, J. Polym. Sci. A-2, 9, 283, 1971.)
Tensile Strength (MPa)
1.0
0.8
0.6
0.4
0.2
–15
–10
–5
0
5
10
Reduced Variable: Ln (aG)
·
a = [( A ) 4.79 · (a ε)]
G
36
T
Figure 7.10 Master curve for the tensile strength of Dow Corning 795 silicone sealant.
Range of applicability: Temperature T = −20°C to +80°C; strain rate: ε˙ = 0.05–250 mm/
min; joint cross-sectional area: 36–420 mm2. Model statistics: coefficient of correlation R =
0.968, standard deviation S = 0.046 MPa, average error = ±6.2%. (From W.S. Gutowski, P.
Lalas, and A.P Cerra, in Science and Technology of Building Seals, Sealants, Glazing and
Waterproofing, Volume 5, ASTM STP 1271, M. Lacasse (Ed.), p. 97, American Society for
Testing and Materials, Philadelphia, PA, 1995.)
9162_C007.indd 207
3/17/09 8:27:32 PM
208
Handbook of Sealant Technology
A disregard of these principles may lead to erroneous results and, consequently,
to wrong conclusions regarding the quality of adhesion. This has been demonstrated
through a comparative study on the adhesion of a structural silicone sealant to a
high-density polyethylene (HDPE) substrate using the following two test protocols:
(1) ASTM C-794 standard [16], and (2) CSIRO protocol involving “thin peel specimens” [11]. The former specifies a 1.6-mm-thick sealant layer in the peel specimens
tested at strain rate [ε˙] of 50 mm/min, while the latter uses a sealant layer thickness
of 0.2 mm.
HDPE, and polyolefins in general, are well known for poor adhesion to adhesives, sealants, or surface coatings. This is a consequence of their chemical inertness arising from the lack of surface chemical groups capable of creating chemical
bonds with adhesives or sealants. Considering these any assessment of the adhesion
strength of untreated HDPE to any sealant should reveal poor adhesion, which, in
turn, should be demonstrated by interfacial delamination of the sealant from the
substrate surface.
Regardless of this anticipation, we have discovered, however, that the joints
made with HDPE substrates and sealed with relatively thick sealant beads (thickness greater than 3 or 6 mm) frequently passed the standard testing involving peel
and tensile specimens. This, in turn, might have inadvertently led to a catastrophic
failure of structural joints involving this particular combination of materials under
real-life service conditions regardless of the fact that the requirements of typical
standards (e.g., ASTM or ISO) were satisfied.
To demonstrate the seriousness of the foregoing problem pointing out the inherent inadequacies of some standards commonly accepted for determining the quality
of sealant adhesion, we carried out experiments involving the following materials
and procedures:
•
•
•
•
Substrate material: HDPE (nonbondable [if not adequately treated] material)
Sealants: structural sealants available from key sealant manufacturers
Joint configuration: peel specimens according to ASTM C-794 standard [16]
Bead thickness: 1.6 mm (as per ASTM C-794 standard requirements) 0.2
mm (CSIRO “thin peel specimen”)
• Peel rates: 0.05 and 50 mm/min.
The photographs in Figure 7.11 illustrate the representative outcome of the foregoing experiments in terms of the appearance of fracture surfaces of peel specimens
involving Dow Corning 795 silicone sealant and HDPE substrate fabricated using
sealant bead thickness of 1.6 mm, as required by ASTM C-794 standard, as well as
0.2 mm (CSIRO “thin specimens”).
It is evident from the photo in Figure 7.11a that the ASTM C-794 “thick peel
specimen” exhibits 100% cohesive fracture within the sealant regardless of the fact
that its interfacial delamination from the HDPE substrate has been anticipated. It can
also be seen from this figure that lowering the peel rate by up to 2 orders of magnitude (from 5 to 0.05 cm/min) was not sufficient to reveal any potential deficiency of
adhesion between the sealant and HDPE.
9162_C007.indd 208
3/17/09 8:27:32 PM
209
Adhesion Testing of Sealants
A
B
0.5
5.0
25
·
Peel Rate: ε (mm/minute)
50
Figure 7.11 Appearance of fracture surfaces of 180° peel specimens prepared using a
structural silicone sealant and HDPE substrate at the following two sealant thicknesses: (a)
1.6 mm (as per ASTM C-794 standard), and (b) 0.2 mm. Specimens were tested using peel
rates of 0.5, 5, 25, and 50 mm/min.
The analysis of the fracture surface of the “thin peel specimen” illustrated in
Figure 7.11b demonstrates, on the other hand, that 100% interfacial delamination
occurs independently of the peel rate, thus confirming inadequate adhesion of the
silicone sealant used in these experiments to HDPE.
The following becomes evident from an analysis of the fracture surfaces in
Figure 7.11:
a. Specimen (a)—thick sealant layer (1.6 mm): The use of a relatively thick
sealant layer between the HDPE substrate and flexible backing tape leads to
a “masking” of the expected poor adhesion by “producing” 100% cohesive
failure within the sealant. This implies that the quality of adhesion satisfies
the requirements of the ASTM C-794 standard, and hence, the sealant could
be inappropriately used in structural applications.
It is apparent that 100% cohesive failure has been observed not only for
the standard-recommended strain rate of 50 mm/min but also for the rate
as low as 0.5 mm/min, that is, 100 times slower. Typically, reducing the
strain rate by two orders of magnitude should allow for the detection of
poor interfacial adhesion through a change in failure mode from cohesive to
100% delamination along the substrate-sealant interface.
b. Specimen (b)—thin sealant layer (0.2 mm): The use of a thin sealant layer
between the HDPE substrate and backing tape creates conditions that promote the disruption of the interfacial bonds before cohesive failure can occur
(see Section 7.5.1 for a detailed discussion of underlying phenomena).
Thus, this test configuration (thin sealant bead layer) yields results that are expected
for a silicone sealant on an untreated HDPE, that is, poor adhesion. Also it is
9162_C007.indd 209
3/17/09 8:27:33 PM
210
Handbook of Sealant Technology
noticeable from the analysis of the fracture surface in Figure 7.11b that even a 100fold increase in the strain rate from 0.5 to 50 mm/min does not cause a change in
failure mode from interfacial to cohesive.
To further demonstrate how the control of the two key factors promoting the
disruption of interfacial bonds, that is, (1) a thin layer of sealant between the
substrates, and (2) low strain (peel) rate applied during specimen peel-off (180°
peel angle), can be effectively used for detecting inadequate sealant–substrate
adhesion, we designed an additional experiment utilizing the following materials
and procedures:
• Sealant bead thickness: (a) “thick” peel specimen”: 1.6 mm (as per ASTM
C-794), (b) “thin peel specimen”: 0.2 mm (CSIRO specification)
• Sealant type: One-component RTV oxime-cured structural silicone
• Substrates: a range of black anodized aluminum substrates (cobalt salt
sealed): see Table 7.2 for details
• Peel rate: 0.05 and 50 mm/min.
• The complete set of data on the failure mode (% interfacial delamination)
and peel force recorded for both “thick” and “thin” peel specimens are provided in Table 7.2 (data originally were reported in Reference 11). Similar
trends have been observed for clear and bronze-colored anodized aluminum [11].
The data in Table 7.2 clearly indicate that the thick (1.6 mm) peel specimens are
not suitable for testing the quality of adhesion between the one-component RTV
oxime-cured structural silicone and a range of anodized aluminum substrates, as this
configuration is insensitive to inherent adhesion problems. However, specimens with
a thin sealant layer provide the desired stress concentration at the sealant–substrate
interface to create adhesion failure (interfacial delamination) relatively independent
of the applied peel rate (0.05 to 50 mm/min).
An analysis of Equations 7.12 and 7.13 in Section 7.5.1.1 provides further insight
into the outcome of the foregoing experiments. It is seen from these equations that
the strength of adhesion at the sealant–substrate interface (or fracture energy G of
the interface) is most accurately determined when the term quantifying plastic dissipation energy is minimized through the reduction of factors Sa and ha in these equations. This is effectively achieved by minimizing the thickness of the sealant layer.
In turn, this provides the desired reduction in deformation energy per unit volume
of sealant due to an increase in the material’s modulus of elasticity with decreasing
thickness (see Section 7.5.1.2). The reduction of peel angle ψ leads to further minimization of the plastic deformation losses.
The foregoing observations, originally reported in our earlier publications [7,11],
were subsequently confirmed by Shephard [17]. He also provided comprehensive
experimental evidence [18] demonstrating a significant increase in the sensitivity of
peel tests carried out at 45° peel angle. This test configuration led to complete interfacial delamination of the sealant that originally exhibited 100% cohesive failure
when tested at identical strain rates but peeled at a standard 180° angle.
9162_C007.indd 210
3/17/09 8:27:33 PM
211
Adhesion Testing of Sealants
Table 7.2
Comparison of the Failure Mode (Percentage of Interfacial Delamination) and
Peel Force for 180° Peel Specimens Comprising a Range of Black-anodized
Aluminum Substrates and a One-component RTV Oxime-cured Structural
Silicone
Thin Specimen: 0.2 mm
Thick Specimen: 1.6 mm
Peel Rate: 50 mm/min
Substratea
25 μm/HS
Peel Rate:
50 mm/min
Percentage of
Percentage of
Peel Force
interfacial
Peel Force
interfacial
(N/mm) Delamination (N/mm) Delamination
3.04
0
1.00
95
Peel Rate:
0.05 mm/min
Peel Force
(N/mm)
Percentage of
interfacial
Delamination
0.24
100
25 μm/CS
1.60
60
0.52
100
0.16
100
25 μm/AS
2.48
0
0.92
60
0.36
100
15 μm/HS
3.24
0
0.80
95
0.18
100
15 μm/CS
3.00
5
0.88
95
0.18
100
15 μm/AS
3.20
0
1.04
95
0.20
100
a
Anodized aluminum (6063 alloy): 15/25 μm anodized layer thickness; HS—hot sealed; CS—cold
sealed; AS—accelerated sealing.
7.7.2 Test Methods based on Tensile Specimens
7.7.2.1 Deficiencies in Existing Test Methods Involving Tensile Specimens
Test methods based on the use of a tensile joint configuration typically employ specimens such as those depicted in Figure 7.7. In most cases the width of the joint, B, is
identical to the glueline thickness G, and typically set at 10 or 12 mm.
As discussed in depth [19–21], most test protocols and standards employing a
tensile test configuration provide poor discrimination in adhesion levels between
sealants and substrates. Typical approaches involve preconditioning the specimens by controlled exposure to static environmental conditions such as (1) heat–
freeze cycling at the extremes of service temperatures [e.g., −29°C and +88°C]
(2) immersion in water, or application of water spray, and (3) UV radiation (xenon
arc, fluorescent UV-A lamps, or carbon arc). Subsequent mechanical testing provides data on the joint strength, elongation at failure, and percent delamination at
the interface.
Mechanical testing, when carried out after artificial exposure, rarely yields significant information on the quality of adhesion except in cases when 100% delamination
occurs. Frequently, partial delamination (e.g., 10%–15%) of sealant is initiated at the
location of the highest stress concentration, that is, along the edge of the extended sealant bead, followed by the crack tip veering off from the interface into the bulk sealant
9162_C007.indd 211
3/17/09 8:27:33 PM
212
Handbook of Sealant Technology
and propagating there with increasing strain. Eventually, the interfacial stress is relieved,
and the joint appears to have failed in a predominantly cohesive failure mode.
As first noted by Cerra [19], standard tensile-based tests can often lead to ambiguous results as follows:
a. A mixed failure mode (partial interfacial delamination and partial cohesive
failure of the sealant) results when the crack path deviates from the interface to the bulk material, masking any poor adhesion due to the resultant
stress relief. This leads to the release of interfacial stresses as the subsequent
strain energy dissipation then continues through the deformation of the bulk
sealant material instead of concentrating at the sealant–substrate interface.
b. Most test protocols for determining sealant adhesion fail to recognize that,
in actual service, joints fail due to the concurrent action of three key degradation factors that adversely affect adhesion, namely,
• Mechanical stress
• Sealant strain
• Moisture present at the interface in the form of a condensed film and/
or water vapor.
The combined action of these factors results in the irreversible disruption of the
interfacial bonds even with silicone sealants, which can usually recover when individual stresses are removed. The presence of moisture together with tensile stress
imposed by the joint strain leads to the rupture of individual bonds, while the resultant crack opening prevents them from being reconstituted, as would have been the
case in absence of strain.
7.7.2.2 CSIRO Test Method for Determining Adhesion of Sealants
7.7.2.2.1 Fundamental Principles of the Method
Molecular bridges, which may initially provide good adhesion across the sealant–substrate interface, can be effectively disrupted under the combined action of mechanical stress, strain, and water molecules. If the strain, energy results in the interfacial
stress concentration at the joint’s edge exceeding the bond’s critical fracture energy,
fracture of the bonds bridging the interface occurs. Under test conditions that apply
continuous and high-enough stress levels and sustained strain, stress relaxation is
prevented, and a sustained crack propagation along the interface takes place.
The principle of applying simultaneous, rather than sequential, mechanical and
hydrothermal stresses to the interface gave rise to the novel test procedure developed by Cerra [19] for determining the strength and durability of adhesion between
elastomeric sealants and rigid substrates. The magnitude of tensile stress applied to
the interface (and combined with immersion in water or exposure to close to 100%
humidity) is well below the cohesive strength of the sealant material. Typically, it
does not exceed 0.138 MPa (20 psi), which is the design stress for structural bonding
applications of sealants in façade engineering.
Another important phenomenon concerning the exposure of elastomeric sealants
to constant stress (as applied in the CSIRO test procedure) has been noted by the
9162_C007.indd 212
3/17/09 8:27:33 PM
213
Adhesion Testing of Sealants
Sealant Bead Thickness (mm)
22
1 to 2 mm
20
18
16
14
12
10
0
1
2
3
4
5
Time (hours)
Figure 7.12 Creep deformation of Dow Corning 795 structural sealant subjected to constant stress under water immersion at 55°C.
authors of this paper and is also relevant to this protocol. Due to their viscoelastic
nature, sealants subjected to appropriate levels of tensile stress will deform through
progressive creep, as illustrated in Figure 7.12 for Dow Corning 795 sealant.
In this example, the material starts deforming through creep after approximately
20 min. The strain rate during this stage of sealant deformation was found to be in
the range of 0.002 to 0.003 mm/min. This is similar to the strain rates that building
façade components experience due to daily and annual temperature fluctuations that
lead to extension–compression cycling of sealants in the facade’s joints.
Based on the foregoing results, it has been demonstrated [12] that the adhesion
and cohesive strengths of the sealants, as determined through the CSIRO test protocol, reflect the material’s properties under similar service conditions that might
occur in curtain walls comprising 2500-mm-long aluminum frame and glass panels
bonded with a structural silicone sealant.
7.7.2.2.2 Technical Details of the Method
The full details of the CSIRO method are described in References 19 and 20. The
quality of sealant–substrate adhesion is determined using tensile specimens (see
Figure 7.7) with the sealant bead dimensions of 12 × 12 × 40 mm. These are subsequently subjected to creep-load conditions under a combination of mechanical and
hydrothermal stresses. This is achieved by exposing the specimens to water immersion
at 50°C and simultaneously applying a predetermined level of stress to the bondline.
Figure 7.13 illustrates the general outline of the CSIRO creep exposure apparatus.
The specimens are inserted into a row of stainless steel clamps at the bottom of
the water baths below the water level.
The broad range of fracture stress is initially determined by increasing the
stress in 69 kPa (10 psi) increments until failure occurs. The permanent load
applied to the sealant–substrate interface is retained for a predetermined period,
9162_C007.indd 213
3/17/09 8:27:34 PM
214
Handbook of Sealant Technology
2
1
3
Figure 7.13 A view of the CSIRO creep apparatus comprising the following key components: (1) water bath preset to desired temperature, for example, 50ºC; (2) system of levers
that transpose the load (3) to tensile specimens, hence applying a predetermined tensile stress
level to the sealant–substrate interface.
for example, 24 h. If the bonded specimen does not fail during this time by delamination along the interface, a new specimen is inserted and the load increased
by the aforementioned incremental value. The cycle, always employing a new
specimen, is repeated up to the stage when the failure stress resulting in specimen
delamination or cohesive failure is achieved. Subsequently, multiple replicates
are tested at gradually increasing stress levels (13.8 kPa [2 psi]).Ultimately, the
accurate value of the true strength of adhesion is determined through the use
of a statistical procedure known as the “Dixon up-and-down” protocol [22]. All
failed specimens are also visually inspected to determine percent delamination
and cohesive failure, and the procedure requires about 15 specimens to achieve a
statistically significant result.
7.7.2.2.3 Comparison of Results from the CSIRO Creep
Test Method and Standard Tensile Tests
A comprehensive analysis of multiple experimental data sets demonstrating the
advantages of the CSIRO test method over currently known test methods is provided
9162_C007.indd 214
3/17/09 8:27:34 PM
215
Adhesion Testing of Sealants
in References 19 and 20. A comparison of typical results is discussed in this section
to demonstrate that the standard tensile test carried out after accelerated exposure
may not discriminate the adhesion properties of sealant–substrate combinations.
Table 7.3 below provides details of substrates bonded in one of our experiments
using a 1-component oxime-cured structural silicone sealant commonly used in façade
engineering for structural bonding of large glass panes to various substrate materials.
Table 7.4, in turn, provides results of adhesion tests carried out after the following
conditioning [denoted (1) to (3)] of the tensile specimens (12 × 12 × 40 mm sealant
bead) fabricated using the one-component oxime-cured structural silicone sealant
and substrates listed in Table 7.3:
Table 7.3
Details of Anodized Aluminum and Glass
Substrates Bonded with a One-component RTV
Oxime-cured Structural Silicone Sealant
Substrate Code
a
Substrate Description
G 10
Float glass with stainless steel coating
B9
Black anodized “accelerated-sealed” Al
EB 3
Black anodized Al (“Chemel” processa)
EA 1
Clear anodized Al (“Chemel” process a)
B9
Black anodized “accelerated-sealed” Al
Anodized (sulfuric acid) and electrolytically colored (using a
tin salt) aluminum.
Table 7.4
Results of Sealant Adhesion Tests in Terms of Tensile Strength Data after
the Following Sample Conditioning Protocols: (1) no Conditioning; (2)
7-day Immersion in 20°C Water; (3) CSIRO Procedure: Concurrent Stress
and 50°C Water Immersion
Conditioning (1)
“Dry Specimen”
Substrate
Code
9162_C007.indd 215
Strength
(MPa)
Interfacial
Delamination
(%)
Conditioning (2)
7-day immersion in
20°C Water
Strength
(MPa)
Interfacial
Delamination
(%)
Conditioning (3)
“CSIRO protocol”
(stress + 50°C water)
Interfacial
Strength Delamination
(MPa)
(%)
G 10
0.85
0
0.85
0
0.65
0
B9
0.83
0
0.80
0
0.52
20
93
EB 3
0.94
0
0.85
0
0.12
EA 1
0.91
0
0.74
0
0.12
92
B9
0.92
0
0.72
0
0.16
78
3/17/09 8:27:34 PM
216
Handbook of Sealant Technology
Table 7.5
Details of Sealants and Substrates Used in Comparison of Various Exposure
Conditions Prior to Testing Sealant Adhesion Using a Tensile Joint
Configuration
Sealant
Sealant
Code
Generic Type
Substrate
Substrate
Code
Substrate Details
A
One-part silicone (alkoxy-cured)
1
G
Two-part polyurethane
7
Anodized Al (dark bronze)
Granite (polished)
H
One-part polyurethane
8
Compressed fiber-cement
G
One-part polyurethane (siliconemodified)
10
Concrete (prepared as per ASTM C719)
1. “Dry” tensile specimens (1 month sealant cure) without any conditioning
2. Specimens immersed in 20°C water for 7 days without stress and strain
3. CSIRO test procedure [see Section 7.7.2.2.2]: simultaneous application of
tensile stress and water immersion (at 50°C).
It is seen from the data presented in Table 7.4 that tensile testing of dry specimens,
or even those immersed (without prestressing) for 7 days in room temperature water
prior to tensile testing always results in 100% cohesive failure of the sealant without
indication of any inherent adhesion problems.
However, as seen from the last column of Table 7.4, the CSIRO test protocol
involving the simultaneous application of hot water immersion, stress, and strain
provides clear discrimination in the quality of adhesion for the sealant–substrate
combinations studied. The results show that only the coated glass displays no adhesion failure.
Another set of experiments involved a range of generic types of structural and
weather sealants and a diversified range of substrates including anodized aluminum, concrete, granite, and compressed fiber-cement sheet as used in building and
construction applications. A description of the individual materials is provided in
Table 7.5, while Table 7.6 provides the comparative test data for tensile specimens
tested as mentioned, that is, dry, after 7-day water immersion at 20°C and using the
CSIRO test procedure.
The results presented in Table 7.6 provide further evidence that the CSIRO test
procedure is better able to discriminate adhesion quality than the more conventional
methods.
For more comprehensive information, readers are referred to complete sets of
experimental data provided in References 19 and 20.
9162_C007.indd 216
3/17/09 8:27:34 PM
217
Adhesion Testing of Sealants
Table 7.6
Results of Sealant Adhesion Tests in Terms of Tensile Strength Data for
the Sealant–Substrate Combinations (Structural and Weather Sealants)
Listed in Table 7.5
Conditioning (1)
“Dry Specimen”
Conditioning (2)
7-day immersion in
20°C Water
Strength
(MPa)
Interfacial
Delamination
(%)
A1
0.62
0
0.82
A7
0.74
0
0.52
G1
0.95
47
0.31
G7
1.00
0
0.47
G8
1.03
0
0.30
G10
1.03
0
0.49
H1
0.41
0
0.36
H7
0.37
0
0.34
H8
0.35
0
H10
0.42
0
J1
0.74
0
0.66
0
0.08
1
J7
0.76
0
0.53
0
0.05
100
J8
0.73
0
0.24
77
0.05
100
J10
0.77
100
0.35
100
0.05
100
Substrate
Code
Strength
(MPa)
Interfacial
Delamination
(%)
12
0.19
25
0
0.35
5
0
0.05
100
30
0.05
100
3
0.13
2
0
0.05
100
17
0.137
20
0
0.11
40
0.33
100
0.14
40
0.33
10
0.95
100
Strength
(MPa)
Interfacial
Delamination
(%)
Conditioning (3)
“CSIRO protocol”
(stress + 50°C water)
Structural Sealant: Silicone
Weather Sealant: Two-part Polyurethane
Weather Sealant: One-part Polyurethane
Weather Sealant: One-part Polyurethane (Silicone-modified)
7.8 Contemporary Standards For
Determining Sealant Performance
The first standards [23–25] for evaluating the performance of elastomeric sealants were
developed for the assessment of weather sealants and focused on the retention of bond
and cohesive integrity rather than load-bearing capacity such as in structural adhesives.
In the mid-1990s, the International Union of Laboratories and Experts in
Construction Materials, Systems and Structures, RILEM (Reunion Internationale
des Laboratoires et Experts des Matériaux, Systèmes de Constructions et Ouvrages)
recognized the deficiencies of the existing protocols for the accelerated evaluation
of the performance of elastomeric sealants, particularly with regard to adhesion
and long-term durability. The task of designated Technical Committee TC139-DBS
9162_C007.indd 217
3/17/09 8:27:35 PM
218
Handbook of Sealant Technology
(Durability of Building Sealants) was to prepare an internationally recognized protocol for the accelerated determination of the performance of elastomeric joint sealants for use in high-movement building facades, which are subjected to continuous
cyclic movements and severe weathering throughout their service life.
7.8.1 RILEM Accelerated Durability Tests
The protocol developed by TC 139-DBS [26] introduces “static” or “dynamic”
conditioning of sealant specimens (50 × 12 × 12 mm) during the 28-day curing
at 23 ± 2°C and 50% ± 5% RH. The dynamic cure simulates the variation of the
working glueline thickness of the sealant bead in the building façade during
cure.
After the completion of 28-day (static or dynamic) cure, the sealant specimens are
exposed to three cycles of accelerated degradation as follows:
a. Artificial weathering exposure
b. Thermomechanical cycling comprising two consecutive cycles of low-temperature extension (−20 ± 2°C) and high-temperature compression (70°C
± 2°C), both carried out at the extremes of the sealant’s rated movement
capability (e.g., ±12.5%, ±20%, or ±25%)
After completion of each degradation cycle (a and b) the specimens are extended to
their rated extension capability and then examined for the following signs of failure:
1. Percent adhesion loss (delamination between sealant and substrate)
2. Loss of cohesion
3. Whether the locus of failure was at the sealant–substrate interface or in the
bulk of the sealant
7.8.1.1 Static Conditioning Protocol
RILEM TC-139 “static conditioning” [26] comprises three cycles of the following
consecutive steps:
1.
2.
3.
4.
3-Day oven exposure at 70°C ± 2°C
1-Day immersion in distilled water at 23°C ± 2°C
2-Day oven exposure at 70°C ± 2°C
1-Day immersion in distilled water at 23°C ± 2°C
7.8.1.2 Dynamic Conditioning Protocol
RILEM TC-139 “dynamic conditioning” [26] recognizes the fact that, in real life,
particularly in the building (e.g., in building facades) or shipbuilding industry, the
joint width varies cyclically after installation due to thermal movement in the substrates. Consequently, the freshly fabricated specimens are inserted into a cyclic
movement device, allowing joint extension or compression at a controlled rate of 70
± 20 mm/min.
9162_C007.indd 218
3/17/09 8:27:35 PM
Adhesion Testing of Sealants
219
The “dynamic conditioning” involves an overall 28-day sealant cure as follows:
a. A total of 14 daily cycles involving the following steps carried out at 23°C
± 2°C and 50°C ± 5 % RH:
1. Extension of the freshly fabricated specimen by 7.5% within 5 min
of fabrication
2. Holding for 2.4 h at 7.5% extension
3. Compression of the specimen by 7.5%
4. Holding for 2.4 h at 7.5% compression
5. Returning the joint to its initial width of 12 mm for the remainder of the
24-h cycle.
b. 14-Day cure in a static state (joint returned to its initial width of 12 mm) at
23°C ± 2°C and 50% ± 5 % RH
7.8.2 RILEM Protocols Involving Artificial Weathering
After the completion of the cure under “static” or “dynamic” conditions, the specimens are subjected to artificial weathering [26–30] involving one of the following:
• Xenon arc (340 nm lamps at 0.5 W/m2 nm) automatic weathering 102 min
of “dry heat” exposure at 65°C ± 2°C and 65% RH
• 18 min of water spray or immersion in water of temperature less than
40°C
• Fluorescent UVA-340/water spray cyclic exposure [28] involving 8-week
cycling as follows:
• 8 h of “dry heat” exposure at 65°C ± 2°C
• 4 h of UV radiation and water spray (water temperature less than 40°C)
7.8.3 Thermomechanical Cycling and Determination
of Sealant Durability
After the artificial weathering outlined in Section 7.8.2, the specimens are subjected to two cycles of thermomechanical stress involving low-temperature extension and high-temperature compression at the sealant’s rated movement capability
as follows:
• Day 1: The specimen is conditioned for 3 h at −20°C ± 2°C, then extended
to its rated movement capability (e.g., +25%) for a period of 21 h.
• Day 2: After release of extension, the specimen is conditioned for 3 h at
70°C ± 2°C, then compressed to its rated movement capability (e.g., −25%)
for a period of 21 h.
• Day 3: Repeat of procedure for Day 1.
• Day 4: Repeat of procedure for Day 2.
• Day 5–7: Release of compression and 3-day storage at 23°C ± 2°C and
50°C ± 5% RH.
9162_C007.indd 219
3/17/09 8:27:35 PM
220
Handbook of Sealant Technology
The specimens are then assessed for
1. Percent adhesion loss (delamination between sealant and substrate)
2. Loss of cohesion
3. Whether the failure occurred at the sealant–substrate interface or in the
bulk of the sealant
The standard RILEM TC-139 protocol [26] requires exposure of the specimens to at
least three cycles of artificial weathering exposure and thermomechanical cycling,
as described earlier. It is recommended, however, that the number of degradation
cycles should be such as to induce substantial (visible) degradation in the least stable
material in the group of candidate sealants.
7.8.4 Most Recent RILEM Recommendations for
Sealant Durability Assessment
The outcome of the most recent work carried out by the RILEM working group
TC 190-SBJ (“Service Life of Building & Construction Joints”) provides further
improvements over the earlier protocols described in Sections 7.8.1 to 7.8.3, by
adopting test and cycling conditions to better reproduce local climatic and service
conditions. The Group, under the direction of Wolf and Enomoto, has recently
prepared a draft document [31] outlining the details of a proposed “RILEM
Technical Recommendation.”
The broad details of the new protocol [31] are as follows:
1. Sealant specimens, 100 × 20 × 15 mm (length × width × thickness) are first
cured and conditioned “statically,” that is, without any joint movement for
28 days at 23°C ± 2°C and 50% ± 5% RH.
2. The fully cured specimens are then exposed to repetitive exposure cycles
(6- week cycles, as described in Section 7.8.2) comprising
• UV light (xenon arc, or fluorescent UVA-340 nm, or carbon arc), heat
and moisture (water spray or complete water immersion)
• Repetitive cyclic movement—manual or automated (6 weeks) carried
out during or after the accelerated weathering period (see earlier) comprising mechanical cycling up to the sealant’s rated movement capability (i.e., ±12.5%, ±25%, or ±50%) with the following cycles:
−− One extension and compression cycle (at a rate of 5.5 ± 0.7 mm/
min) per 24 h for 4 consecutive days
−− Specimen at rest (no joint movement) for 3 days
3. After each exposure cycle (item 2 in this list) the specimens are extended to
their full rated movement capability (i.e., +12.5%, +25%, or +50%), and the
joint is examined for signs of failure.
4. This procedure is repeated as often as required to achieve a visible degree
of degradation.
9162_C007.indd 220
3/17/09 8:27:35 PM
Adhesion Testing of Sealants
221
It is noteworthy that the proposed RILEM TC 190-SBJ protocol broadly resembles
the procedure adopted by the ASTM C 1519 standard [25], which recommends the
following test procedure:
• Sealant specimen (12.5 × 12.5 × 50 mm) cure for 21 days under the following conditions: (a) 7 days at 23°C ± 2°C and 50% ± 5% RH; (b) 7 days at
38°C/95% RH; and (c) 7 days at 23°C ± 2°C and 50% ± 5% RH.
• A 4-week artificial weathering (xenon arc or Fluorescent UVA-340 nm/
water spray cyclic exposure: 2 h UV light/2 h water spray), followed by 6
cycles of extension and compression at room temperature.
• Joint examination for signs of failure (including quantification).
• Repeat cycles until some degradation is visible.
7.9 Summary
This chapter provides a comprehensive analysis of the theoretical and practical aspects of the principal factors and mechanisms controlling interfacial failure
between sealants and substrates. The factors influencing the properties and in-service performance of elastomeric sealants are analyzed with emphasis on their shortand long-term adhesion.
An in-depth review and analysis of contemporary test methods for determining
the adhesion behavior of sealants indicated a number of deficiencies in the existing test protocols and standards. It is shown through numerous practical examples
that, in order to demonstrate satisfactory strength and durability of adhesion, one
needs to subject the interface to a set of conditions likely to promote disruption of
interfacial bonds before the cohesive failure of bulk sealant occurs. It is shown that,
to achieve this, it is necessary to simultaneously apply both mechanical and accelerated environmental stress, the former at the levels below the cohesive strength of
sealant material.
A practical guide on reliable test protocols designed to ascertain satisfactory longterm performance of sealants in service is also provided.
References
1. American Society for Testing and Materials (ASTM) Standard No. C1401-02: “Standard
guide for structural glazing” (2002).
2. American Society for Testing and Materials (ASTM) Standard No C1193-05: “Standard
guide for use of joint sealants” (2005).
3. D.H. Kaelble, Trans. Soc. Rheology, 4, 45 (1960).
4. D.H. Kaelble, Trans. Soc. Rheology, 3, 161 (1959).
5. D.H. Kaelble, Trans. Soc. Rheology, 9, 135 (1965).
6. A.N. Gent and G.R. Hamed, J. Appl. Polym. Sci, 21, 2817 (1977).
7. W.S. Gutowski and E.R. Pankevicius, Fatigue Fract. Eng. Mater Struct., 17, 351 (1994).
8. K. Kendall, J. Phys. D.: Appl. Phys., 8, 1449 (1975).
9. S. Wu, Polymer Interface and Adhesion, p. 535, Marcel Dekker, New York (1982).
10. E.J. Ripling, S. Mostovoy, and H.T. Corten, J. Adhesion, 3, 107 (1971).
9162_C007.indd 221
3/17/09 8:27:35 PM
222
Handbook of Sealant Technology
11. W.S. Gutowski, L. Russell, and A. Cerra, in Science and Technology of Building
Seals, Sealants, Glazing and Waterproofing, Volume 2, ASTM STP 1200, J.M.
Klosowski (Ed.), p. 87, American Society for Testing and Materials, Philadelphia,
PA (1992).
12. W.S. Gutowski, P. Lalas, and A.P Cerra, in Science and Technology of Building Seals,
Sealants, Glazing and Waterproofing, Volume 5, ASTM STP 1271, M. Lacasse (Ed.), p.
97, American Society for Testing and Materials, Philadelphia, PA (1995).
13. J.D. Ferry, J. Am. Chem. Soc., 72, 3746 (1950).
14. M.L. Williams, R.F. Landel, and J.D. Ferry, J. Am. Chem. Soc., 77, 3701 (1955).
15. A.N. Gent, J. Polym. Sci. A-2, 9, 283 (1971).
16. American Society for Testing and Materials (ASTM) Standard No C794-06: “Standard
guide for structural glazing” (2006).
17. N.E. Shephard, in Durability of Building Sealants, RILEM Proc., Volume 37, A.T. Wolf
(Ed.), p.161, E&FN Spon, London (1999).
18. N.E. Shephard and J.P. Wightman, in Science and Technology of Building Seals,
Sealants, Glazing and Waterproofing, Volume 5, ASTM STP 1271, M. Lacasse (Ed.), p.
226, American Society for Testing and Materials, Philadelphia, PA (1995).
19. A.P. Cerra, J. Testing Eval., 23, 370 (1995).
20. A.P. Cerra and W.S. Gutowski, in Science and Technology of Building Seals, Sealants,
Glazing and Waterproofing, Volume 5, ASTM STP 1271, M. Lacasse (Ed.), p. 209,
American Society for Testing and Materials, Philadelphia, PA (1995).
21. J. Iker, and A.T. Wolf, in Proceedings of the Symposium on Building Sealants: Materials,
Properties and Performance, ASTM STP 1062, T.F. O’Connor (Ed), p. 67, American
Society for Testing and Materials, Philadelphia, PA (1990).
22. W.J. Dixon, J. Am. Stat. Assoc., 69, 967 (1965).
23. American Society for Testing and Materials (ASTM) Standard No. C794-80, “Standard
test method for adhesion-in-peel of elastomeric joint sealants” (1980).
24. American Society for Testing and Materials (ASTM) Standard No. C719-05, “Adhesion
and cohesion of elastomeric joint sealants under cyclic movement (Hockman cycle)”
(2005).
25. American Society for Testing and Materials (ASTM) Standard No. C1519-04,
“Evaluating durability of building construction sealants by laboratory accelerated
weathering procedures” (2004).
26. RILEM Standard: RILEM TC 139-DBS: “Durability Test Method—Determination of
changes in adhesion, cohesion and appearance of elastic weatherproofing sealants for
high movement façade joints after exposure to artificial weathering,” Mater. Struct., 34,
579 (2001).
27. ISO 4892-Part 2: 2006 Plastics, “Method of exposure to laboratory light sources—Part
2: Xenon lamps” (2006).
28. ISO 4892-Part 3: 2006 Plastics, “Method of exposure to laboratory light sources—Part
3: Fluorescent UV lamps” (2006).
29. (a) ISO 9047: 1989, and (b) ISO 9047: 2001. “Building Construction—Jointing
Products—Determination of adhesion/cohesion properties of sealants at variable temperature” (1989 and 2001).
30. ISO/DIS 11600: 2002, “Building Construction—Jointing Products—Classification and
Requirements for Sealants” (2002).
31. A.T. Wolf, and N. Enomoto, “Durability Test Method—Determination of changes in
adhesion, cohesion and appearance of elastic weatherproofing sealants after exposure of
statically cured specimens to artificial weathering and mechanical cycling,” in RILEM
Technical Committee 190-SBJ: Document No. NO15, Proposed RILEM Technical
Recommendation (RTR) Accelerated Weathering Test Method—Work Group Draft
Version 2006-08031 (2006).
9162_C007.indd
222
View publication stats
3/17/09 8:27:36 PM